Cooperative Interference Management in Multi-Antenna Wireless Networks

다국어 초록 (Multilingual Abstract)

Due to the rapidly increasing demand of higher data rate with a large number of connected devices, interference has become the key factor in designing future wireless networks, such as the fifth generation (5G) cellular network. This is mainly because the more nearby wireless devices transmit at the same time, more dominant interference causes the signal distortion, thereby limiting the overall system performance. Motivated by this, interference has been deemed an intrinsically destructive phenomenon so that it must be avoided as much as possible. In this dissertation, I reevaluate the fundamental notion of interference in wireless networks, and introduce a new paradigm that interference can be a friend in many wireless networks, rather than a foe in the prevailing traditional paradigm. To support this, I explore and discuss the new opportunities of harnessing interference to improve the system performance for a wide range of multi-antenna wireless networks, especially under the imperfect knowledge of channel state information at transmitter (CSIT).
To begin with, I focus on interfering multiple access channel as a canonical model of uplink cellular networks. Under the assumption of delayed CSIT, I demonstrate that harnessing overheard interference signals at the receivers as side information brings significant performance gains. For a certain network configuration, it is shown that the proposed scheme achieves the optimal degrees-of-freedom (DoF) of the uplink cellular network. The same spirit of exploiting interference can be extended into multi-hop interference networks beyond single-hop network. I demonstrate that relays can be used to provide DoF gains in two-hop interference networks through the proposed cooperative relaying techniques even when imperfect CSI is available at relays. Under two different CSI assumptions at relays, I also characterize the feasibility conditions on relays’ configuration to accomplish the target DoF gains. A major implication is that the cooperative relay can help to steer the transmitters’ signals in a way that it judiciously exploits the past receptions to reconstruct the overheard interference at each receiver as side information for future transmissions, instead of the transmitters.
Lastly, I turn my attention to the new interference issues, especially for key radio access technologies of 5G cellular systems: non-orthogonal multiple access (NOMA), full-duplex (FD) radios, and wireless powered communications network (WPCN). I shed light on the various potential benefits of harnessing interference via coordination between nodes. In particular, it is shown that interference can be exploited as a new source of energy harvesting to enable self-sustainable wireless networks as well as a means of increasing number of multiplexed users for the scenario in which massive connectivity is required over limited radio resources, such as massive machine-type communication (mMTC). The finding has a special importance in designing 5G new radio (NR) systems, showing that such a cooperation between nodes can help overcome the limiting effects of interference.

목차 (Table of Contents)

• 1 Introduction 1
• 1.1 Background: Interference in Wireless Networks 1
• 1.1.1 Single-Hop Interference Networks 3
• 1.1.2 Multi-Hop Interference Networks 6
• 1.2 Contributions and Organization 9
• 1.3 Notation 11
• 2 Retrospective Interference Alignment for Uplink Cellular Networks 12
• 2.1 Introduction 13
• 2.2 System Model 17
• 2.3 Retrospective IA using outdated CSIT 18
• 2.3.1 Achievable Scheme for (M,N,K) = (2, 2, 2) 19
• 2.3.2 Achievable Scheme for (M,N,K) = (K, 2,K) 25
• 2.3.3 General Achievability 30
• 2.4 Discussion: Analysis of sum-DoF Gain 39
• 2.4.1 Sum-DoF Gain from Outdated CSIT 39
• 2.4.2 Comparison with Two-User MIMO-IC 40
• 2.4.3 A Sum-DoF Outer Bound 41
• 2.5 Conclusion 44
• 3 Relay-Aided Interference Networks with Imperfect Channel Knowledge – Part I: Hybrid CSI at Relay 46
• 3.1 Introduction 47
• 3.2 System Model 51
• 3.2.1 A MIMO Relay-Aided MISO Interference Channel 51
• 3.2.2 A MIMO Relay-Aided SISO X-Channel 54
• 3.2.3 Partial Channel Knowledge Assumption 55
• 3.3 Relay-Aided Space-Time Beamforming 56
• 3.3.1 Phase 1: Side-Information Learning 57
• 3.3.2 Phase 2: Space-Time Relay Transmission 58
• 3.4 Multiplexing Gain of The K-user MISO Interference Channel With a MIMO Relay 61
• 3.5 Multiplexing Gain of the K ⇥ L SISO X-Channel With a MIMO Relay 69
• 3.6 Discussion: Assessment of Practical Issues 75
• 3.6.1 CSIT Acquisition Overhead Analysis 75
• 3.6.2 Validity of Delayed CSI Assumption 76
• 3.7 Conclusion 77
• 4 Relay-Aided Interference Networks with Imperfect Channel Knowledge – Part II: Delayed CSI at Relay 79
• 4.1 Introduction 80
• 4.2 System Model 82
• 4.3 Relay-Aided Retrospective Interference Alignment 84
• 4.3.1 Motivating Example (MT = MR = K = 2) 84
• 4.3.2 Sum-DoF Analysis for the General Case 87
• 4.3.3 Extension to Layered Two-hopK-user MISO IC with a MIMO Relay 90
• 4.4 Discussion: Sum-DoF Gain Analysis from MIMO relay with Delayed CSI 93
• 4.5 Conclusion 95
• 5 Non-Orthogonal Multiple Access in Multi-Cell Networks: Theory, Performance, and Practical Challenges 96
• 5.1 Introduction 97
• 5.1.1 Background 97
• 5.1.2 Theory Behind NOMA 99
• 5.1.3 Related Work: A Review of Recent Advance 104
• 5.1.4 Contribution 107
• 5.2 System Model 108
• 5.3 New CBF Design for Multi-Cell MIMO-NOMA 109
• 5.3.1 Algorithm I: ICA-CBF NOMA 110
• 5.3.2 Algorithm II: IA-CBF NOMA 112
• 5.3.3 Discussion 114
• 5.4 Numerical Results 116
• 5.5 Practical Challenges for Multi-Cell NOMA 119
• 5.5.1 SIC Implementation Issues 119
• 5.5.2 Imperfect CSI 120
• 5.5.3 Multi-User Power Allocation and Clustering 120
• 5.5.4 Operation with FFR 121
• 5.5.5 Security 122
• 5.6 Conclusion 122
• 6 In-Band Full-DuplexWireless: An Interference Management Perspective 124
• 6.1 Introduction 125
• 6.2 System Model 128
• 6.2.1 FD-BS-HD-User Cellular Networks 128
• 6.2.2 FD-BS-FD-User Cellular Networks 131
• 6.2.3 Degrees of Freedom 132
• 6.3 Cyclic Interference Alignment 132
• 6.3.1 Symmetric FD MIMO Cellular Networks 133
• 6.3.2 Asymmetric FD MIMO Cellular Networks 141
• 6.4 Extension to Multi-Beam Cases 147
• 6.5 (M,N,K) FD-BS-FD-user Cellular Networks 152
• 6.6 Numerical Results and Discussions 155
• 6.6.1 Achievable Ergodic Sum Rate Performance 155
• 6.6.2 Performance in Multi-cell Networks 158
• 6.7 Conclusion 164
• 7 Embracing Interference in Wireless Powered Communication Networks: Interference Harvesting 166
• 7.1 Introduction 167
• 7.2 System Model 169
• 7.3 A Weighted Sum-Rate Maximization 172
• 7.4 A Total Transmission Time Minimization 174
• 7.5 Simulation Results 178
• 7.6 Conclusion 181
• 8 Conclusion 183
• 8.1 Summary 183
• 8.2 Future Research Directions 185